基于PROSAIL模型及TM与实测数据的MODISLAI精度评价  

Assessment of MODIS LAI product accuracy based on the PROSAIL model,TM and field measurements

作　　者：杨飞[1,2] 孙九林[1] 张柏[2] 姚作芳[2,3] 王宗明[2] 王卷乐[1] 乐夏芳[1] 

机构地区：[1]中国科学院地理科学与资源研究所资源与环境信息系统国家重点实验室,北京100101 [2]中国科学院东北地理与农业生态研究所,长春130012 [3]中国科学院研究生院,北京100089

出　　处：《农业工程学报》2010年第4期192-197,I0001,共7页Transactions of the Chinese Society of Agricultural Engineering

基　　金：国家自然科学基金项目(40871187);国家自然科学基金项目(40771146)等共同资助

摘　　要：利用PROSAIL模型和TM影像反演的叶面积指数LAI及地面实测数据,从空间和时间序列上分别对吉林省主要玉米种植带的MODISLAI产品精度进行验证分析,为准确评价和科学准确应用MODIS数据提供科学依据。研究结果表明,MODISLAI产品精度因玉米生长时期和植被类型而异,混合像元是造成MODISLAI产品低估玉米作物LAI值的主要原因之一;不同年份不同生长时期,MODISLAI和实测数据表现出不同的关系;在玉米生长灌浆期之后,MODISLAI产品与同期地面实测玉米LAI相比,约低估了33%～53%左右,MODISLAI产品值明显低于TM影像反演的LAI值,约低估玉米LAI值约30%～69%左右。因MODISLAI产品值与实际值有较大偏差,非常有必要对农作物MODISLAI产品值进行验证分析,研究结果可为准确利用MODISLAI产品及其算法改进后产品精度提供一种评价途径。By comparing the MODIS LAI to field measurement and retrieved LAI from the PROSAIL physical model using TM image,accuracy of the MODIS LAI product for corn crops in Jilin Province was assessed,which could provide the basis of evaluating and using MODIS products accurately.The results show that the accuracy of the MODIS LAI product varies largely with the crop growth stages and the vegetation types.In addition,the mixed pixels were likely causing underestimation of the MODIS product for corn crops.The differences between MODIS and field measured LAI varies with different years and growth periods.After grain-filling stage,MODIS LAI values were apparently lower than estimated LAI by TM image and PROSAIL physical model.The LAI of corn crops had been underestimated by MODIS about 33%–53%and about 30%–69%,as compared with the field measurements and simulated LAI from TM, respectively.As a whole,MODIS LAI product of corn crop has much bias compared with actual LAI value,it is necessary for the validation of corn crop MODIS LAI product accuracy.This study can provide the scientific reference for the use of MODIS LAI product,and it can also be used as an approach for validating the MODIS LAI product generated by improved algorithms.

关 键 词：模型 作物 图像质量 叶面积指数 MODIS TM 精度验证 

分 类 号：TP79[自动化与计算机技术—检测技术与自动化装置] S127[自动化与计算机技术—控制科学与工程]